Targeting treatable traits in severe asthma: a randomised controlled trial

Discussion

People with severe asthma require improved management strategies [6, 8, 25]. We report the results of a personalised medicine approach to severe asthma where we show that multidimensional assessment can identify the many treatable traits that exist in this disease, and that targeting treatments to these traits leads to significant improvements in HRQoL, asthma control and asthma attacks. To our knowledge this is the first RCT to test a multidimensional and targeted approach to severe asthma, and the first treatable-traits RCT to be reported. This work advances knowledge of personalised disease-management strategies for this challenging airway disease, and provides the much needed data to support recently proposed new treatment paradigms [5, 11, 13, 14, 26].

There have been many calls to rethink the approach to complex airway diseases [5, 11, 13, 14, 25, 27]; however, until now there have been no RCTs evaluating the efficacy of a multidimensional assessment followed by personalised targeted treatments. A meta-analysis of severe asthma multidimensional assessment showed improved asthma control, HRQoL and reduced exacerbations and recommended that RCTs were needed [16]. Similar conclusions were reached by the Lancet commission [5], and others [26]. The application of personalised medicine in airway diseases can be achieved using an approach that carefully evaluates the treatable traits [13] and applies therapy that is precisely tailored to the specific disease traits that manifest in each individual [13, 17]. Clinical trials of inflammometry in asthma and COPD emerged in the early 2000s, where individuals with eosinophilic airway inflammation were identified and treatment implemented to target this inflammatory endotype, leading to reduced exacerbations [28–30]. Monoclonal antibody therapies targeting type-2 high inflammation in asthma have further progressed this area [31]. These studies support a personalised-medicine paradigm where inflammatory endotypes are identified and treatment is targeted to specific pathways [26]. We applied this approach to the eosinophilic inflammatory endotype, and additionally included macrolide antibiotics to target the neutrophilic endotype [32, 33] and statins for systemic inflammation [17], in addition to targeting comorbidities and risk factors. It is important to note that our study predated the availability of anti-interleukin-5 therapy in Australia. As such, we used oral corticosteroids (OCS) to target eosinophilic airway inflammation. We recognise the need to adapt our algorithm to include currently available and effective treatments that target eosinophilic disease without the toxicity of OCS [34]. However, the use of OCS provides important knowledge in terms of proof of concept for this trial.

Additionally, we recognise that personalised medicine extends beyond one pathway [26], and is an approach where treatments are targeted to the needs of individuals, not only on the basis of biomarkers, but also genetic, phenotypic or psychosocial characteristics that distinguish the individual [35]. Hence, approaches to complex airway diseases need to consider these factors, especially since individuals with severe asthma express multiple treatable traits [26, 36]. This RCT reports the success of this holistic approach. In our study, quality of life improved to a greater degree than that seen in recent studies of monoclonal antibodies and macrolide antibiotics [34], suggesting that the multidimensional intervention which targets traits in each domain is multiplicative, and treatment beyond pharmacotherapy is necessary. We propose that treating the multiple manifestations has an additional stacked effect and this is one of the major benefits of the treatable-traits approach.

The patient-reported outcome of HRQoL was the primary end-point. The impact of targeted treatment on this outcome was large. This is consistent with a proof-of-concept controlled study of a similar intervention in COPD [17]. In that study SGRQ was the primary outcome and improved by 14 units; the SGRQ in our current study improved by a similar magnitude of 13.4 units, supporting the reproducibility of the intervention and the results. Severe asthma has a major impact on health status that results in long-term and devastating burdens [6, 8, 25]. Therefore, improving health status is an important goal from a patient perspective, and which has not yet been achieved despite advances in pharmacotherapy [37]. This was an appropriate outcome for this study, as the measure assesses multiple domains, which was necessary for the multicomponent intervention tested. Interestingly and surprisingly, we did not observe a change in health status in the control group, which involved management in an established severe asthma clinic. Several factors may have influenced this. First, the usual-care group did not include a case manager to assist with implementation of the treatment plan; this likely resulted in the control participants receiving fewer treatments over the course of the intervention period. Despite this, the usual-care group were treated by an expert multidisciplinary team. Second, the intervention was delivered as a simultaneous approach compared to the sequential approach of the severe asthma clinic [38]; as such the intensity of the treatment in the intervention was greater. However, this supports the approach proposed, as the right patients receive the right treatments within an efficient time period. Finally, the intervention involved an algorithm for treatment of inflammatory endotypes, which all intervention participants received, as opposed to a more ad hoc approach in the control group [18]. We advocate that the magnitude of the effect is achieved by not one individual aspect of the programme, but the “stacking” or additive effects of multiple treatment interventions [17].

Asthma attacks are another important outcome in severe asthma [36, 39]. We demonstrated significantly fewer unscheduled primary-care presentations for acute asthma in the intervention group. For other acute asthma events there were trends toward fewer attacks. Our follow-up period was only 4 months, which was probably not long enough to evaluate the impact of the intervention on exacerbations, and nor was the study powered for this outcome. Nonetheless, these results are promising, and should be addressed in future studies.

We report the results of a multidimensional assessment in a large group of severe asthma patients. We show that multidimensional assessment can be used to identify the prevalence of these treatable traits, and demonstrate the significant trait burden experienced by severe asthma patients. Furthermore, the number of traits were significantly correlated with health status and each additional trait identified was associated with a clinically significant decrement in SGRQ, a result we have also reported in COPD [15]. In the absence of a multidimensional assessment many of these important clinical characteristics may remain unrecognised, and therefore untreated, as is reported in a systematic review of severe asthma assessment [16]. The traits we included in the assessment and individualised interventions were comprehensive, but not exhaustive. Identifying traits that have the biggest impact on outcomes whether that be exacerbations or quality of life is important for the design of future treatable trait trials and for implementation into practice [11]. We have previously reported traits that predict future exacerbation in severe asthma, but more work is needed in this area [36].

Another important finding from a disease-management and self-efficacy perspective, was that there were suboptimal self-management traits in patients with severe asthma. This suggests that the basic management strategies and goals are not adequately achieved as part of usual disease management. Suboptimal self-management skills are associated with increased risk of attacks in severe asthma [36], and our intervention ensured the assessment of these traits and applied individualised treatments to improve them. An important component was the use of a case manager in the delivery of the intervention. This was a key difference from the usual-care severe asthma clinic [18], and suggests the importance of this role in the delivery of complex and multicomponent interventions [39].

Our RCT was a complex intervention and provides novel insight for the design of future studies examining multidimensional assessment and individualised management. Our approach was comprehensive and targeted all of the traits identified. The intervention group did receive more treatment visits than the control group. We did not seek to define how many visits the control group received, as we sought to replicate usual care within a severe asthma clinic. Furthermore, we sought to perform a pragmatic RCT, comparing our intervention to current best practice. However, we do acknowledge that this could potentially impact on the results. Nevertheless, we believe that this is an important consideration for future implementation studies. It is likely that this complex population will require more intensive intervention than current practice allows. This is an important observation that needs to be recognised. We propose that healthcare systems need to re-evaluate care for individuals with high healthcare needs [14]. Yet, whether an approach that included fewer traits would have led to the same effect is not known, but is worth considering.

While the provision of this intervention resulted in improved outcomes for patients with severe asthma, it recognisably comes with an increased economic cost. A health-economic evaluation is required in future work. We hypothesise that the delivery of personalised medicine would become cost-neutral or even result in cost savings by reducing the iatrogenic consequences of toxic treatment and the cost of asthma attacks and burden on the acute healthcare sector. Economic modelling suggests that the direct healthcare cost savings that would flow from better targeted treatment for asthma in Australia ranges from AUD 129 million to AUD 1288 million [40].

Our study has a number of strengths and some limitations. In terms of our RCT, the sample size is relatively small, and the intervention and control groups were not matched for contact time. This means that larger trials are needed in order to inform practice and implementation of such an approach. Still, we urge the importance of these data particularly when evaluating such complex and multicomponent interventions [41], and therefore argue the necessity of these data in progressing this important area of airway disease management. Furthermore, the follow-up time was only 4 months; future trials should include a longer follow-up to assess the impact of interventions that require significant behaviour change, for example adherence, diet and physical activity and of course the evaluation of exacerbation of time.